How to Move Crypto From an Exchange to Your Wallet
Moving crypto from an exchange to your own wallet safely comes down to four checks done in order: copy the receive address straight from your wallet and confirm it on the wallet's own screen, pick the withdrawal network your wallet actually supports, send a small test amount first, and only then send the rest. A transfer between wallets you own is not a taxable sale under IRS guidance, but a blockchain payment cannot be called back, so the checks have to happen before you press confirm, not after.
That last point is the reason this page exists. Bitcoin.org puts it plainly: a Bitcoin transaction cannot be reversed, it can only be refunded by the person receiving the funds. The FTC's consumer guidance says the same about crypto payments in general. An exchange withdrawal is the moment your coins leave a system with a support desk and enter one with none. The routine below is built around that handover: every step is something you can verify while the exchange still holds the coins. Copying an address from your transaction history is the habit address poisoning scams exploit; see address poisoning explained.
What You Need Open Before You Start
Have two screens in front of you, and do not start the withdrawal until both are ready.
The wallet side. Open the wallet that will receive the coins and go to its receive screen for the specific asset. Note three things: the address it shows, the network or chain it says that address is for, and whether the wallet supports the token you plan to send on that network. A wallet that holds bitcoin on the Bitcoin network does not automatically hold a wrapped bitcoin token on an Ethereum layer 2, and a wallet that holds ether may or may not display a stablecoin on a chain it has not been told about. If you are new to how wallets hold keys rather than coins, our explainer on what a crypto wallet does covers the model in plain terms.
The exchange side. Open the withdrawal screen for the same asset and look at the network dropdown before entering anything. Many assets can be withdrawn on several networks. Write down which networks the exchange offers, the fee shown for each, and the minimum withdrawal. You are looking for the overlap between what the exchange can send and what your wallet can receive.
Your recovery backup. If the receiving wallet is new, confirm you have written down and checked its recovery phrase before any value lands in it. A wallet you cannot restore is a worse place for coins than the exchange you are leaving. The wallet recovery seed checklist walks through testing a backup without exposing it.
Only when all three are in place does it make sense to move value.
Step One: Take the Address From the Wallet and Nowhere Else
The receive address must come from the wallet that will hold the coins, generated fresh from its receive screen. Not from an old email to yourself, not from a note on your phone, and not from your transaction history.
That last source is the dangerous one. Trezor's support documentation on address poisoning describes the attack: scammers create a vanity address that closely resembles one you have used, then send zero-value transactions or worthless imitation tokens so that the lookalike address appears in your own history. The bet is that you will later copy an address from that history and pick the wrong line. Trezor says these attacks mainly target Ethereum and other EVM chains, where low fees make mass attacks cheap, and its guidance is to avoid copying addresses from your wallet app, from block explorers, or from your past transactions, including when moving funds off an exchange.
The second threat sits between your clipboard and the paste. Trezor's explanation of its trusted display lists three ways an on-screen address check can fail on an ordinary computer or phone: a fake or tampered app shows an address that is not yours, clipboard malware swaps the address after you copy it, or a plain copy-paste error. A hardware wallet's own screen is the defence for the first two. When you open a receive screen in the wallet's companion app, the device can show the same address on its screen, and you compare the two. If they differ, something on your computer changed what you were shown. That screen is only trustworthy if the device itself is, which is why we wrote a separate guide on how to tell whether a hardware wallet was tampered with.
Trezor's documentation is honest about the limit of that check: the device can confirm that nothing changed along the way, but it cannot tell you whether an address someone gave you is the right one. For a withdrawal to your own wallet that limit does not bite, because you are the one who generated the address.
A practical comparison routine for the paste step:
For Bitcoin specifically, the address format itself tells you something. An address starting with bc1q is native SegWit, bc1p is Taproot, 3 is a script hash and 1 is legacy. If the exchange offers several and your wallet produced a bc1q address, that is the one to paste. Our guide to Bitcoin address types explains the prefixes and which wallets can pay which.
Step Two: Match the Network on Both Ends
The single most common way people lose coins on a withdrawal is choosing a network the receiving wallet does not support. The address can look valid while the destination cannot use it.
The reason is structural. On Ethereum and the chains that copied its account design, one private key produces the same 0x address on every such chain. Trezor's network guide gives the example that Ethereum and Avalanche C-Chain addresses are identical in its app even though they are separate networks, and it warns that because EVM networks share an address format, an address looking valid on several chains does not confirm which network you are on. The chains are kept apart by a chain ID built into each transaction, introduced in EIP-155 so that a transaction valid on one chain cannot be replayed on another; Ethereum mainnet's chain ID is 1. That separation is what makes a wrong-network withdrawal a real problem: the coins land at your address on a chain your wallet may not be watching, or on a chain the receiving service does not support at all.
Trezor's guidance on exchange withdrawals contains one line worth repeating: do not simply choose the cheapest option when withdrawing. The cheapest network in the dropdown is often a layer 2 or a different chain entirely, and it is only a good choice if your wallet supports that asset on that network. If the layer 1 versus layer 2 distinction is new, our layer 1 vs layer 2 guide explains what changes when a token lives on a rollup.
A three-question check before you select a network:
The Failure Table: What Goes Wrong and Whether It Can Be Fixed
The table below sorts the common withdrawal mistakes by what actually happened on-chain and who, if anyone, can help. It is built from the protocol documentation and wallet guidance cited in the sources section; outcomes for any specific case depend on the services involved, and nothing here promises a recovery.
| Mistake | What happened on-chain | Who controls the coins now | Usually fixable? | First move |
|---|---|---|---|---|
| Right network, address belongs to someone else (typo or poisoned lookalike) | A valid transfer to a stranger's address | The stranger | Almost never; the recipient must choose to return it | Stop, document the transaction ID, report if it was a scam |
| EVM token sent on a different EVM chain to your own self-custody address | Coins sit at your address on the other chain | You, via the same key | Often, if your wallet can add that network | Add the network in a wallet that supports it and check the balance |
| Asset sent on a network your hardware wallet's app does not support | Coins at an address derived from your key on an unsupported chain | You in principle, but your tools may not reach them | Sometimes, with other software; Trezor warns it can mean permanent loss | Contact the wallet maker's support; do not hand your seed to anyone offering help |
| Missing or wrong destination tag or memo on a deposit to an exchange | Coins arrive at the exchange's shared address with no customer reference | The receiving exchange | Often, through that exchange's support, at its discretion | Open a ticket with the transaction ID and the amount |
| Token sent to an exchange on a network the exchange does not support for that asset | Coins at the exchange's address on a chain it is not crediting | The receiving exchange | Varies by exchange; some decline | Ask the receiving exchange's support; expect a fee or a refusal |
| Wrong asset entirely (for example, a token sent to an address meant for a different coin) | Depends on whether the address formats are compatible | Varies | Rarely straightforward | Ask the receiving service before anything else |
Two patterns stand out. First, mistakes that land coins at an address you control, on any chain, are usually recoverable with the right software, because the key is still yours. Second, mistakes that land coins with a third party depend on that third party's willingness, and a stranger has no obligation at all. That asymmetry is why the address check in step one matters more than anything else on this page.
The table also shows why recovery offers deserve suspicion. Anyone who contacts you claiming they can reverse a transfer is describing something the blockchain does not allow.
Memos and Destination Tags: When You Need One
Some networks attach a small reference field to a payment. Exchanges use it to pool many customers behind one address and still tell deposits apart.
On the XRP Ledger this field is the destination tag. The XRPL documentation describes destination tags as indicating the beneficiary or destination for a payment, formatted as 32-bit unsigned integers. They have no effect on the ledger itself; they tell an off-ledger system, such as an exchange, which customer to credit. When a tag is missing, the documentation notes, it is not immediately obvious which customer to credit, and resolving it can take manual work. An account holder can switch on a RequireDest setting so the ledger rejects payments that lack a tag instead of accepting money nobody can attribute.
On Stellar the equivalent is the memo. Stellar's developer documentation describes it as an optional unstructured data field, with types including a text memo of up to 28 bytes and an ID memo that is a 64-bit unsigned integer. It also notes that memos were previously used to tell apart individual accounts in a pooled account, a role that muxed accounts now serve.
What this means for a withdrawal to your own wallet:
The direction matters. Leaving a tag off a deposit into a pooled exchange address is the classic error, and the fix runs through the receiving exchange's support desk.
Step Three: The Test Transaction Routine
A test transfer costs one extra network fee and buys certainty that the address, network and wallet all line up. Trezor's network guide recommends it explicitly and adds two practical details: when withdrawing from an exchange, use the exchange's minimum withdrawal as the test amount, and only send the full amount after the test funds appear in the receiving wallet.
The routine, in order:
A worked example with stated inputs
Suppose an exchange account holds 0.5 BTC and the owner wants all of it in a hardware wallet. The numbers below are hypothetical, chosen only to show the arithmetic; real minimums and fees are set by each exchange and change.
Test transfer: 0.0005 BTC, fee 0.00005 BTC. Balance left on the exchange: 0.5 minus 0.00055, which is 0.49945 BTC.
Main transfer: the exchange's "max" button typically deducts the fee, so the owner sends 0.49940 BTC and pays 0.00005 BTC.
Total received in the wallet: 0.0005 plus 0.49940, which is 0.4999 BTC. Total fees: 0.0001 BTC, of which 0.00005 BTC is the price of the test.
Whether that price is worth paying depends on the size of the main transfer. On a 0.5 BTC move, the hypothetical test fee is one hundredth of one percent of the amount. On a transfer that is itself only a few times the minimum, a test may not make sense, and the extra care shifts to the address comparison instead.
Two cautions about this example. Exchanges set their own withdrawal fees, which can differ from the network fee they actually pay. And on Ethereum and other account-based chains, a test transfer of a token also costs gas on the main transfer, so the test is paid twice there too, as Trezor's address-poisoning guidance points out.
Step Four: Allowlisting, Confirmation Emails and Holds
Many exchanges let you save withdrawal addresses and some let you restrict withdrawals to that saved list. The idea is simple: an attacker who gets into your account cannot withdraw to their own address without first adding it, and adding it triggers a confirmation step and often a delay.
Kraken's help article on adding and confirming a new withdrawal address shows how one exchange implements this. A new address must be added with a unique label before it can be used. Addresses added from a trusted device are approved automatically; otherwise approval goes through a link sent to the account's registered email, and that link is valid for one hour. The same article says that if the account password is changed without sign-in two-factor authentication or a master key in place, withdrawals to new addresses are held for 24 hours for security purposes, while addresses already saved are not affected. Kraken also warns that different cryptocurrencies may share the same address format, so the address type should be confirmed when saving.
Other exchanges use different names (address book, allowlist, whitelist) and different delays. The points that carry across:
For the wider question of how much to keep on any exchange and how to judge a platform's reserves, our exchange reserve risk checklist covers what proof-of-reserves reports do and do not show.
Reading the Fee Before You Confirm
There is no point ranking exchange fees here: they change, they differ by network, and the number on a summary page is only a guide. Kraken's own fee page says withdrawal fees are subject to change and that the final fee is shown at confirmation. What helps is a method for comparing the options in front of you on the day.
| Check | Where to find it | Why it matters |
|---|---|---|
| Fee for each available network | Withdrawal screen, network dropdown | The same asset can cost very different amounts on different networks |
| Is the fee flat or variable? | Exchange fee page and the confirmation summary | A flat fee makes larger, less frequent withdrawals cheaper per coin |
| Minimum withdrawal | Withdrawal screen | Sets the floor for a test transfer |
| Amount received after fees | Confirmation summary | The only number that tells you what lands in the wallet |
| Does your wallet support this network? | Wallet documentation | A cheap network your wallet cannot see is not cheap |
| Cost to move the coins later | Network fee conditions on the receiving chain | Some low-fee networks are costly to bridge back from |
The comparison is a two-line calculation for each candidate network: amount received equals amount withdrawn minus the exchange fee, and the fee as a share of the transfer equals the fee divided by the amount withdrawn. Run it for the networks your wallet supports and ignore the rest.
On Bitcoin, the network fee paid on-chain depends on mempool conditions, while an exchange's withdrawal fee is a charge the exchange sets to cover its transaction costs, so the two numbers need not match. Our mempool and transaction fee explainer shows how fee rates in satoshis per virtual byte translate into dollars.
When the Exchange Asks Where the Coins Are Going
A withdrawal screen may ask whether the destination is a wallet you own, an account at another exchange, or someone else's wallet, and sometimes for the name of the recipient or the receiving platform. These questions come from anti-money-laundering rules that require regulated platforms to collect and pass along certain information about transfers, commonly called the Travel Rule. How much is asked, and above what amount, depends on the country.
For a withdrawal to your own hardware or software wallet, answer that the wallet is self-hosted and belongs to you. Some platforms then ask you to prove ownership, for example by signing a message with the wallet or by sending a small amount back. Answer the questions accurately; a wrong answer can delay or block the withdrawal. Our sibling page on the crypto Travel Rule explains the thresholds and why self-hosted wallets draw these questions.
Watching It Arrive
Once the withdrawal leaves the exchange, three stages follow, and confusing them causes needless worry.
Pending at the exchange. The exchange may run security or compliance checks, or apply a hold, before broadcasting. No transaction ID exists yet.
Broadcast, unconfirmed. A transaction ID appears in your withdrawal history. You can look it up on a block explorer for that network. It is waiting to be included in a block.
Confirmed. The transaction is in a block, and each new block on top adds a confirmation. Bitcoin.org notes that transactions accumulate confirmations, each making them increasingly difficult to reverse. Your wallet may show the incoming amount while it is unconfirmed; treat it as settled only after it confirms.
If the exchange shows "complete" but your wallet shows nothing, look the transaction ID up on an explorer before anything else. The explorer answers three questions: which network it went out on, which address received it, and whether it confirmed. If the network is not the one your wallet is watching, see the failure table above. If the address does not match yours, stop and treat it as a possible compromise of your device or account. Our field-by-field guide to reading a transaction on a block explorer shows where each of those answers sits on the page.
Taxes: Moving Coins to Yourself Is Not a Sale
For U.S. federal taxes, the IRS digital assets page addresses this directly in its guidance on the yes-or-no digital asset question on Form 1040. Transferring digital assets from one wallet or account you own or control to another you own or control is listed as a situation where you can answer "No". The exception is a transfer fee paid in digital assets: the IRS page says that if you paid a transaction fee with digital assets, that is a digital asset transaction, and it lists paying a transfer fee with digital assets among the dispositions that require a "Yes" answer.
The IRS frequently asked questions on virtual currency, which the IRS labels as generally applying to transactions completed before January 1, 2025, give the same answer for the transfer itself: moving virtual currency between wallets, accounts or addresses you own is a non-taxable event, even if a platform sends you an information return about it.
What does carry forward is your record-keeping burden. Once coins leave an exchange, that exchange no longer tracks them, and a later exchange you deposit into will not know what you originally paid. The IRS FAQ on identifying specific units refers to records showing when each unit was acquired, its basis and its fair market value at acquisition. Keep the original purchase history and the withdrawal record, including the transaction ID, together. For how this plays out on the broker form that now reports sales, see our explainer on Form 1099-DA, and for the wider rules, the U.S. crypto tax guide. This is general information about the IRS position, not tax advice, and rules outside the United States differ.
Edge Cases and Common Mistakes
Withdrawing a token without the gas coin. On Ethereum and similar chains, a wallet holding only a token cannot move it later without the chain's native coin to pay fees. Plan to hold a small amount of the gas asset on the same network.
Treating a wrapped asset as the original. A bitcoin-denominated token on an Ethereum layer 2 is not bitcoin on the Bitcoin network. It carries the risks of the issuer or bridge behind it, and your Bitcoin wallet will not show it.
Reusing an exchange deposit address as a personal address. An exchange deposit address belongs to the exchange. If you close the account, coins sent there later go to the exchange, not to you.
Sending to a wallet you have not restored from backup. Until you know the recovery phrase works, a hardware wallet is a single point of failure. Test the backup first.
Confirming on autopilot. The withdrawal summary screen is the last place a wrong network or address is cheap to fix. Read every line.
Trusting someone who offers to help. Messages that appear after a failed transfer offering to recover funds, or "support agents" who ask for your recovery phrase, are a known scam pattern. No legitimate exchange or wallet maker needs your seed phrase. If you are worried about what your wallet has already signed, our guide to token approvals and how to revoke them explains a different but related risk.
Rushing a large move during network congestion. High fees and slow confirmations are uncomfortable but not dangerous. If the exchange lets you wait, waiting is fine; if it does not, the transfer will still confirm.
After the Coins Arrive
When the full balance is in your wallet, do three things. Confirm the balance on the device or in the wallet app against the explorer. Save the withdrawal record and transaction IDs with your purchase records. And decide where the recovery phrase lives long term, because from now on there is no exchange account to reset.
If you have not yet written a recovery phrase plan that would survive a fire, a move or a family member needing access, the next page to read is our recovery seed checklist, which turns that decision into a short list of storage and testing steps. More guides in this series are collected on the self-custody and security hub.
Frequently Asked Questions
How do I transfer crypto from an exchange to my own wallet safely?
Open your wallet's receive screen and copy the address it generates, confirming it on the device screen if you use a hardware wallet. On the exchange, choose a withdrawal network your wallet supports for that asset, send the exchange's minimum as a test, check that it arrived using the transaction ID on a block explorer, and then send the rest to the same saved address.
What happens if I withdraw crypto on the wrong network?
It depends on where the coins landed. If they went to your own address on another EVM-compatible chain, the same key usually controls them and adding that network in a supporting wallet often shows the balance. If they went to a service that does not support that network, recovery is up to that service. Funds sent to someone else's address are generally gone unless the recipient returns them.
Do I need a memo or destination tag when sending to my own wallet?
Normally no. Tags and memos exist so an exchange can tell customers apart behind one shared address. A self-custody address belongs only to you, so the tag field is usually left blank. When sending to an exchange account, a tag shown on its deposit screen is required.
Is sending a test transaction worth the extra fee?
For a large transfer, usually yes, because one extra fee is small next to the cost of a wrong address or network. Wallet maker Trezor recommends sending the exchange's minimum withdrawal first and waiting for it to appear before sending the full amount. For a transfer only a few times the minimum, careful address checking may be the better trade-off.
Is moving crypto from an exchange to my own wallet taxable?
Under IRS guidance, a transfer between wallets or accounts you own or control is not a taxable sale, and you can answer "No" to the Form 1040 digital asset question for that transfer alone. Paying a transfer fee in digital assets is treated as a digital asset transaction. Keep your purchase records, because the next platform will not know your cost basis.
Why is my exchange holding my withdrawal or asking who owns the wallet?
Holds often follow security changes such as a new password or a newly added address; Kraken, for example, documents a 24-hour hold on new withdrawal addresses after certain password changes. Questions about who owns the destination wallet come from anti-money-laundering transfer rules, and answering accurately keeps the withdrawal moving.
How long does a withdrawal from an exchange take to arrive?
There are two clocks: the exchange's processing time before it broadcasts, and the network's confirmation time after. Once a transaction ID appears, you can track the second part on a block explorer. A transaction that has been broadcast but not confirmed is still waiting for a block, not lost.
Sources
Wallet makers and exchanges are cited for how their own products document these steps, not as recommendations. The worked example uses hypothetical fees and minimums.
CryptosEyes publishes general educational research, not investment, legal or tax advice. Exchange procedures, fees and holds change; read your exchange's and wallet's current documentation before moving funds, and speak to a qualified tax professional about your own situation.
Source & Review Basis
This article is reviewed against the source types below. Source links are provided to help readers verify primary documents, market context, and methodology independently.
Own-wallet transfers answered No on the digital asset question; transfer fees paid in digital assets are a digital asset transaction. Checked October 22, 2026.
Q38 non-taxable transfers between your own wallets; Q40 record requirements. FAQs generally apply to pre-2025 transactions. Checked October 22, 2026.
Email approval link valid one hour; 24-hour hold on new addresses after certain password changes. Checked October 22, 2026.
Fees subject to change; final fee shown at confirmation. Checked October 22, 2026.
Destination tag purpose, 32-bit format, RequireDest. Checked October 22, 2026.
Memo types and limits. Checked October 22, 2026.